Biochemistry 1982, 21, 3750-3757
3750
peptide, therefore, must be by dissociation of the peptide and rebinding by attachment of its amino-terminal end to the binding sites on the protein (VII). Pepstatin binds tightly (Marciniszyn et al., 1976a) to those regions of the binding site (P3, P,, P1) that had accommodated those residues that became the carboxyl-terminal end of the peptide and can do so as soon as the sites are vacated, even while the peptide is still weakly held to the enzyme at points of contact remote from the catalytic site (Va). This provides an alternative route by which the peptide may dissociate from the protein (V to VIa). In the case of pig pepsinogen activation, that alternative route is apparently faster than the one available in the absence of pepstatin (V to VI). The first chicken peptide liberated is longer than the pig peptide, and apparently its more extensive contact with pseudopepsin holds it more effectively to the distal site(s) on the protein, so that, in the presence of pepstatin, the chicken peptide is released more slowly than is the pig peptide. References Al-Janabi, J., Hartsuck, J. A., & Tang, J. (1971) J . Biol. Chem. 247, 4628-4632. Bohak, Z. (1969) J . Biol. Chem. 244, 4638-4648. Bohak, Z. (1973) Eur. J . Biochem. 32, 547-554. Christensen, K. A., Pedersen, V. B., & Foltmann, B. (1977) FEBS Lett. 76, 214-218. Dunn, B. M., Deyrup, C., Moesching, W. G., Gilbert, W. A., Nolan, R. J., & Trach, M. L. (1978) J . Biol. Chem. 253, 7269-7275.
Dykes, C. W. (1978) Ph.D. Thesis, University of Wales. Dykes, C. W., & Kay, J. (1976) Biochem. J . 153, 141-144. Gray, W. R. (1967) Methods Enzymol. 11, 139-151. Harboe, M., Andersen, P. M., Foltmann, B., Kay, J., & Kassell, B. (1974) J . Biol. Chem. 249, 4487-4494. Harish-Kumar, P. M., & Kassell, B. (1977) Biochemistry 16, 3846-3849. Kay, J., & Dykes, C. W. (1976) Biochem. J . 157,499-502. Kay, J., & Dykes, C. W. (1977) in Acid Proteinases: Structure, Function and Biology (Tang, J. J. N., Ed.) pp 103-130, Plenum Press, New York. Keilova, H., Kostka, V., & Kay, J. (1977) Biochem. J. 167, 855-858. Kivelson, D. (1960) J . Chem. Phys. 33, 1094-1 106. Kostka, V., Keilova, H., & Baudys, M. (1981) in Proteinases and Their Inhibitors (Turk, V., & Vitale, L., Eds.) pp 125-140, Pergamon Press, Oxford. Marciniszyn, J., Jr., Hartsuck, J. A., & Tang, J. (1 976a) J . Biol. Chem. 251, 7088-7094. Marciniszyn, J., Jr., Huang, J. S., Hartsuck, J. A., & Tang, J. (197613) J . Biol. Chem. 251, 7095-7102. McPhie, P. (1976) Anal. Biochem. 73, 258-261. Pechere, J.-F., Dixon, G. H., Maybury, R. H., & Neurath, H. (1958) J . Biol. Chem. 233, 1364-1372. Pedersen, V. B., Christensen, K. A., & Foltmann, B. (1979) Eur. J. Biochem. 94, 573-580. Twining, S. S., Sealy, R. C., & Glick, D. M. (1981) Biochemistry 20, 1267-1272.
Stepwise Sequence Determination from the Carboxyl Terminus of Peptided Joseph L. Meuth,' David E. Harris, Francis E. Dwulet,f Mary L. Crowl-Powers, and Frank R. N . Gurd*
ABSTRACT:
The thiocyanate method for stepwise degradation of peptides from their COOH termini [Stark, G. R. (1968) Biochemistry 7, 17961 has been investigated. The method involves first the reaction of the COOH-terminal residue with thiocyanate in an activation solvent of acetic acid and acetic anhydride and then cleavage of the COOH-terminal residue as its 2-thiohydantoin by acetohydroxamate in aqueous solution. The two steps of the degradation have been studied by using model peptides, and conditions have been developed for the rapid efficient removal and identification of the COOHterminal residue of short peptides. The methods have been applied to peptides that have been covalently attached to insoluble supports. In this solid phase version of the degradation, a highly substituted porous glass activated with N,N'-
carbonyldiimidazole has been prepared for use as the insoluble support. A number of peptides have been coupled to the porous glass, and several rounds of the degradation have been performed on immobilized peptides. High-pressure liquid chromatography provides a rapid, sensitive identification method for the 2-thiohydantoins. In addition, gas-liquid chromatography of the amino acid 2-thiohydantoins and reconversion to the parent amino acid have been used to identify the cleaved residues. The method of sequential degradation has been applied to a number of short model peptides such as Gly-Leu-Tyr, Met-enkephalin, and Val-Leu-Ser-Glu-Gly and has been used to determine the COOH-terminal sequence of 4 residues of a 22-residue cyanogen bromide fragment of pygmy sperm whale myoglobin.
s c h l a c k & Kumpf (1926) proposed a chemical method for the sequential degradation of a peptide from its COOH ter-
minus. The method involved the activation of the COOHterminal carboxyl group by the formation of a mixed anhydride with acetic acid and then reaction with ammonium thiocyanate to form a peptidyl 2-thiohydantoin. The 2-thiohydantoin was cleaved from the peptide by treatment with base to expose a new COOH-terminal amino acid. Stark (1968) reduced the severity of most of the reaction conditions and used the degradation for subtractive sequencing and then later (Cromwell & Stark, 1969) reported methods for direct identification of the cleaved thiohydantoins by thin-layer chromatography (TLC)' or by reconversion to the amino acids (Stark, 1972).
t From the Department of Chemistry, Indiana University, Bloomington, Indiana 47405. Received March 9, 1982. This is the 130th paper in a series dealing with coordination complexes and catalytic properties of proteins and related substances. For the preceding paper, see March et al. (1982). This work was supported by US.Public Health Service Research Grants HL-05556 and HL-14680. *Present address: Department of Molecular Immunology, Scripps Clinic and Research Foundation, La Jolla, CA 92037. Present address: Indiana University Medical Center, Clinical Long, Indianapolis, IN 46223.
0006-2960/82/0421-3750$01.25/0 I
,
0 1982 American Chemical Societv
COOH-TERMINAL PEPTIDE SEQUENCING
Yamashita (1971) described studies on the cleavage of peptidyl thiohydantoins and advocated the use of the acidic form of a cation-exchange resin for mild hydrolysis. The sequence analysis of up to about ten amino acids from the COOH termini of polypeptides has been reported without details (Yamashita & Ishikawa, 1972). The use of thiocyanic acid rather than thiocyanate salts has resulted in improvements in the coupling step (Kubo et al., 1971). The greater reactivity of thiocyanic acid compared to thiocyanate salts has been shown to facilitate the formation of COOH-terminal2-thiohydantoinswith a number of proteins and protein fragments (Dwulet & Gurd, 1979). Developments in methods for the identification of amino acid thiohydantoins include gas-liquid chromatography (GLC) (Rangarajan et al., 1973; Dwulet & Gurd, 1977), two-dimensional TLC (Rangarajan & Darbre, 1975), and highpressure liquid chromatography (HPLC) (Schlesinger et al., 1979; this work). A number of workers have reported solid-phase versions of the degradation. The attachment of peptides to insoluble supports facilitates the recovery of the peptide after each degradation cycle and shortens the time required for each cycle. Porous glass beads and modified polymers have been used as supports [Williams & Kassell, 1975; Rangarajan & Darbre, 1976; Darbre & Rangarajan, 1975; see also Darbre (1977)]. The present work describes the use of carbonyldiimidazole-activated porous glass as the support for the covalent immobilization of a number of peptides in high yield, the stepwise COOH-terminal sequence analysis of immobilized peptides using thiocyanic acid for the coupling step, and the use of HPLC for the identification and quantitation of the cleaved amino acid 2-thiohydantoins. Experimental Procedures
Materials Acetic acid (Mallinckrodt, reagent grade) was redistilled before use. Acetic anhydride (Fisher) was mixed with activated charcoal and redistilled. Acetone (Mallinckrodt, nanograde) was dried over magnesium sulfate. Ammonium thiocyanate (Mallinckrodt, analytical reagent grade) was recrystallized from absolute methanol and stored desiccated. Dimethylformamide (Mallinckrodt, reagent grade) was redistilled from benzene (Stewart & Young, 1969). Triethylamine (Mallinckrodt) was redistilled from ninhydrin before use. All the above reagents were stored at -20 OC. Benzylamine, 2-hydroxy- l-naphthaldehyde, and acetohydroxamic acid (Aldrich) were used without purification. N,N'Carbonyldiimidazole (Aldrich) was stored desiccated at -20 "C. (7-Aminopropy1)triethoxysilane and succinic anhydride were obtained from Pierce. Controlled-pore glass (CPG 10-75 A, 200-400 mesh) was obtained from Electro-Nucleonics, Inc. The amino acid 2-thiohydantoins were prepared as described (Dwulet,& Gurd, 1977). Gly-L-Leu was obtained from Vega. Gly-DL-Leu-DL-Ala, Gly-L-Leu-L-Tyr, and DL-Leu-DL-Phe were obtained from Sigma. Val-Leu-Ser-Glu-Gly was obtained from Cyclo. Met-enkephalin was obtained from Calbiochem-Behring. Crystalline porcine glucagon (lot no. 258I Abbreviations: TLC, thin-layer chromatography; GLC, gas-liquid chromatography; HPLC, high-pressure liquid chromatography; Abu, a-aminobutyric acid; Nle, norleucine; CM-Cys, carboxymethylcysteine; Kb CB-1, NI#j-terminal cyanogen bromide fragment of the major component myoglobin of pygmy sperm whale (Kogia breuiceps) representing residues 1-55; Kb CB-3, COOH-terminal cyanogen bromide fragment of the major component myoglobin of pygmy sperm whale representing residues 132-153.
VOL. 21, NO. 16, 1982
3751
D30-138-2) was provided through the courtesy of Eli Lilly and Co. Major component myoglobin from frozen muscle tissue of pygmy sperm whale (Kogia breviceps) supplied by Drs. R. Bonde and K. Beck of the National Fish and Wildlife Laboratory, Gainesville, FL, was prepared by the method of Hapner et ai. (1968). The peptides Kb CB-1 and Kb CB-3 were prepared by cyanogen bromide cleavage of pygmy sperm whale myoglobin and were purified by gel filtration by methods essentially identical with those used in the preparation of the cyanogen bromide fragments of the myoglobin of Amazon river dolphin (Dwulet et al., 1975). All other reagents were of the highest available quality. The Altex glass (3 mm X 150 mm) microbore chromatography column and glass column jacket were obtained from Rainin Instrument Co., Inc.
Methods Preparation of Thiocyanic Acid. The method is based on that of Stokes & Caine (1907) and resembles method I11 of Dwulet & Gurd (1979).2 Preparation of Derivatized Glass. The controlled-pore glass was reacted with (y-aminopropy1)triethoxysilaneby a method based on that described by Robinson et al. (197 1). The amino content of the aminoalkylsilyl glass was determined by the method of Schmitt & Walker (1977). The aminoalkylsilyl glass was succinylated by a method based on that of Venter & Dixon (1974).2 The extent of the succinylation was determined indirectly by quantitating the content of amino groups. Activation of Succinyl Glass with Carbonyldiimidazole. The activation of the carboxyl groups was performed by a reaction suggested by Staab (1962). One gram of succinyl glass (having between 0.3 and 0.6 mmol of carboxyl groupfg of glass) was suspended in 6 mL of anhydrous dimethylformamide, and to the suspension was added solid N,N'carbonyldiimidazole to a 20-25-fold molar excess over carboxyl groups. The suspension was degassed and was shaken at room temperature for 20 h. The glass was washed with dimethylformamide and with dichloromethane before being dried under vacuum. The activated glass was stable for at least 18 months when stored desiccated at -20 OC. Studies on COOH- Terminal Thiohydantoin Formation. The effect of the composition of the coupling solution on the extent of peptidyl 2-thiohydantoin formation was examined by dissolving 1 pmol of Gly-L-Leu in 1.2 mL of a coupling solution. The reaction mixture was stirred at 5 5 OC for 90 min. At the end of this time, 50-pL samples were removed, dried, and hydrolyzed for amino acid analysis. The decrease in the Leu content of the peptide [determined by amino acid analysis after hydrolysis of the peptide in the absence of 2mercaptoethanol as described by Dwulet & Gurd (1977)l was taken as a measure of the extent of thiohydantoin formation. Preparation of Acetyl- Phe-Leu- 2- t hiohydantoin Using Ammonium Thiocyanate. The compound acetyl-Phe-Leu-2thiohydantoin used in studies on thiohydantoin cleavage was prepared by a method based on the procedure of Yamashita (197 1).2 Studies on COOH-Terminal Thiohydantoin Cleavage. The rate of thiohydantoin cleavage was determined spectroscopically. To 3 mL of cleaving reagent in a quartz spectrophotometer cell (1-cm path length) was added 60 pL of acetylPhe-Leu-2-thiohydantoin solution (containing 165 nmol of peptide). After the solution was mixed, the decrease in absorbance at 290 nm of the solution was measured with a Cary 14 spectrophotometer.
* See paragraph at end of paper regarding supplementary material.
3752 B I O C H E M IS T R Y
MEUTH ET AL.
Table I: COOH-Terminal Solid-Phase Degradation step 1 2 3 4 5 6
soln
volume (mL)
coupling SOPsystem volume (-2 mL) + 1 mLb water wash 4 cleaving sohd 4 water wash 4 activation s o h e 4 coupling soh system volume
time (min) recycledC 90 single pass single pass single pass single pass single pass
17 17 17 17
9
1.5 M HSCN in acetone-activation solution (1:4 v/v). See text. Pumping the coupling solution (23 mL) through the column in a single pass gave no improvement in either the efficiency of removal of the carboxyl-terminal residue or the recovery of the 0.2 M acetohydroxamate in 0.5 M sodium cleaved residue. phosphate buffer, pH 8.3. e Acetic anhydride-acetic acid (4: 1 v/v).
COOH- Terminal Solution Degradation. In a typical experiment, 800 nmol of trifluoroacetic acid treated G~Y-LLeu-L-Tyr was dried in a screw cap tube. The peptide was redissolved in 1.2 mL of acetic acidacetic anhydride (1:2 v/v), and the solution was stirred in a water bath at 55 OC for 5 min. A 0.21-mL sample of 1.4 M thiocyanic acid in acetone was added (300 pmol of HSCN), and the reaction was stirred at 55 O C for 90 min. The solution was dried under a nitrogen stream and then under vacuum. To the dried peptide was added 1 mL of cleaving solution (Table I), the tube was flushed with nitrogen, and the solution was stirred at 55 OC for 30 min. The solution was allowed to cool to room temperature and was extracted 3 times with 1.2 mL of ethyl acetate. The extracts were combined and were dried under a nitrogen stream. The thiohydantoin was identified and quantitated by HPLC and by amino acid analysis after reconversion to the parent amino acid (Dwulet & Gurd, 1977). Covalent Attachment of Peptide to Activated Glass. In a typical experiment, 800 nmol of Gly-L-Leu-L-Tyr was dissolved in 1 mL of anhydrous trifluoroacetic acid. The solution was left at room temperature for 5 min, and the trifluoroacetic acid was then removed by a stream of nitrogen. The dried peptide was washed twice with 1 mL of dichloromethane which was removed by a stream of nitrogen and finally dried under vacuum. The peptide was redissolved in 1 mL of dimethylformamide, and 60 pL of triethylamine was added. The solution was added to 100 mg of activated glass in a screw cap tube, and the suspension was degassed. The tube was flushed with nitrogen, and the reaction was shaken at room temperature for 20 h. At the end of this time, the glass was recovered by filtration and was washed with dimethylformamide, water, saturated Na2HP04,water, 1 M acetic acid, water, and finally with acetone. The peptideglass was dried under vacuum, and the amount of peptide attached was determined by acid hydrolysis and the amino acid analysis of 10-20 mg. In a number of experiments, after the 20-h peptide attachment the excess activated carboxyl groups were blocked by the addition of 0.1 mL of ethanolamine/ 100 mg of activated glass. The reaction was shaken at room temperature for 2 h, and the coupled glass was recovered and washed as described above. COOH- Terminal Solid-Phase Degradation. The degradation was performed with the peptideglass packed in a 3-mm diameter Altex microbore column modified for the use of plunger ends. The column was fitted with a water jacket and was maintained at 60 "C throughout the degradation. The column was flowed upward at 14 mL/h by using a Sage Instruments Model 375 A tubing pump. The peptide-glass was suspended in activation solution (Table I) for packing in the
column (40 mg of peptide-glass occupied approximately 1 cm of the column). The column was washed with 2 mL of coupling reagent (approximately one column volume), and the degradation was then performed as shown in Table I. At the end of step 6, the column was set up to recycle the coupling solution, and the degradative cycle (steps 1-6) was then repeated. The coupling solution was pumped from a reservoir (kept in an ice bath) containing 1 mL of coupling solution, and the column effluent was returned to the reservoir. All other solutions were flowed through the column in a single pass. All solutions were degassed before use. The column effluent was collected in 1-mL fractions which were analyzed as described below, after changing to the cleaving solution (step 3). Identification and Quantitation of Amino Acid 2- Thiohydantoins. The GLC of amino acid thiohydantoins and reconversions to the amino acids were performed as described (Dwulet & Gurd, 1977). The HPLC analyses were performed with a Varian Model 5000 chromatograph fitted with a 4-mm diameter X 30 cm MicroPak MCH-10 analytical column and a Varian Universal guard column filled with Vydac reversephase packing. The column effluent was passed through a Varichrom variable wavelength detector, and thiohydantoins were detected by their absorbance at 265 nm. The solvents used for the analyses were prepared by dilution of a stock solution of 40 mM sodium acetate buffer, pH 3.7; 2.72 g of sodium acetate trihydrate was dissolved in 400 mL of glassdistilled water, and the pH of the solution was lowered to 3.7 by the addition of glacial acetic acid. The solution was made up to 500 mL with glass-distilled water and was then filtered through a MF-Millipore 0.45-pm filter (Millipore Corp). Solvent A was prepared by diluting the stock buffer solution with 3 volumes of glass-distilled water, and solvent B was prepared by mixing stock buffer solution (100 mL), glassdistilled water (60 mL), and acetonitrile (240 mL). The column was flowed with solvent A at room temperature at 3 mL/min (typical operating pressure was 180 atm), and the amino acid thiohydantoins were eluted by a gradient program consisting of 0% solvent B for 5 min after injection and then a 30-min linear gradient to 100% solvent B. Standard solutions of 2-thiohydantoins were made up in methanol. The exact concentrations of the thiohydantoins were determined spectroscopically by dilution of the stock solutions with methanol. The extinction coefficient at 262 nm for all samples except tryptophan and dehydrothreonine was taken to be 17 500 (Cromwell & Stark, 1969). The cleaving solutions that were analyzed by HPLC were injected directly onto the column. For analysis by GLC or by reconversion, the solutions were extracted 3 times with 1.5 volumes of ethyl acetate. The ethyl acetate extracts were combined and dried by a stream of nitrogen. The dried residues were analyzed as described (Dwulet & Gurd, 1977). Results HPLC of Amino Acid 2- Thiohydantoins. The separation of amino acid 2-thiohydantoins by reverse-phase HPLC using gradient elution is shown in Figure 1 and Table 11. Most of the 2-thiohydantoins are well resolved. The thiohydantoins that are poorly separated (dehydrothreonine and valine, lysine and methionine) are resolved by isocratic elution (Table 111). HPLC provides a rapid and sensitive method for the identification and quantitation of amino acid 2-thiohydantoins. Thiohydantoin Formation in Solution. The extent of thiohydantoin formation was studied with the model peptide Gly-L-Leu. Amino acid 2-thiohydantoins are stable to acid hydrolysis at 110 OC in the absence of reducing agents
COOH-TERMINAL PEPTIDE SEQUENCING
Table 11: HPLC of Amino Acid 2-ThiohydantoinP amino acid 2-thiohydantoin
amino elution acid 2-thiotime (rnin)b hydantoin Val LY s Met TYr Ile Leu Nle Phe TIP
1.36 (8) 1.64 (8) 2.18 (5) 2.35 (8) 2.75 (8) 3.15 (7) 3.32 (4) 6.3 (3) 8.8 (3) 9.7 (11)
ASP G~Y Gh CM-Cys Glu Ala His Abu '4% ThI
elution time (min)b 9.9 (12) 11.0 (13) 11.6 (12) 11.8 (9) 13.9 (10) 14.8 (11) 15.5 (10) 16.2 (9) 24.8 (9)
Over the course of approximately 9 months, the elution times of the amino acid 2-thiohydantoins were observed to shorten (by as much as 0.5 rnin for those amino acid 2-thiohydantoins with elution times greater than 10 min). Reliable identification by The numHPLC was achieved by the routine use of standards. ber of determinations is given in parentheses. a
Table 111: HPLC of Amino Acid Thiohydantoins by Isocratic Elution elution time (rnin)
"
amino acid thiohydantoin
gradient elution"
isocratic elutionb
Abu Thr Val LY s Met Tyr
6.3 9.7 9.9 11.0 11.6 11.8
2.5 3.8 4.3 3.9 5.3 5.3
See text. solvent A.
The elution was performed at 3 mL/min with 90%
/
2o0. I
Elution Time
Imin)
FIGURE1: Separation of amino acid 2-thiohydantoins by reverse-phase H P L C on a 4-mm diameter X 30 cm MicroPak MCH-10 column as described in the text. The sample size was 0.5-1.0 nmol of each 2-thiohydantoin. The amino acid 2-thiohydantoins are identified by a single letter representing the parent amino acid: (1) CM-Cys; (2) Abu; (3) Nle.
(Dwulet, 1976), and so the decrease in leucine content of the peptide by amino acid analysis was taken as a measure of thiohydantoin formation. The extent of thiohydantoin formation by a number of coupling solutions at 55 OC is shown in Table IV. The presence of the trifluoroacetic acid or of
VOL. 21, NO. 16, 1982
3753
Table IV: Effect of the Coupling Solution on Thiohydantoin Formation %thiohydantoin formationb
coupling solnu acetic acid-acetic anhydride-trifluoroacetic acid anhydride (33:50:17 v/v) acetic acid-acetic anhydride (1:1 v/v) acetic acid-acetic anhydride (1 :2 v/v) acetic acid-acetic anhydride (1:3 v/v) acetic acid-acetic anhydride (1:4 v/v) acetic acid-acetic anhydride (1:9 v/v) trifluoroacetic acid-acetic anhydride (1 :9 v/v)
76 72 87 100 100 100 57
" The coupling solutions contained in a total volume of 1.2 mL The ex1 pmol of Gly-L-Leu and 300 pmol of thiocyanic acid. tent of Leu-2-thiohydantoin formation in the peptide Gly-L-Leu after 90 min at 55 "C was determined as described in the text. Table V: Rate of Cleavage of Acetyl-Phe-Leu-2-thiohydantoinat Room Temperature
cleavage reagent
half-time of Leu-2-thiohydantoin release
12 M HC1 5.7 M HC1 saturated triethylamine 50 mM acetohydroxamate, pH 8.0a 100 mM acetohydroxamate, pH 8.0" 200 mM acetohydroxamate. PH 8.0a
>30 min >30 min 42 s 90 s 40 s 30 s
a Solutions of acetohydroxamic acid in water were adjusted to pH 8.0 with 5 M NaOH.
trifluoroacetic acid anhydride reduces the coupling efficiency. Trifluoroacetic acid [used in thiohydantoin formation by Kubo et al. (1971)l and trifluoroacetic acid anhydride were used to improve the solubility of proteins and large peptides in the coupling solution at 38 O C (Dwulet & Gurd, 1979). During the course of this work, it was observed that several small to moderate sized peptides were soluble at 55 OC in activating solutions composed of only acetic acid and acetic anhydride. The addition of trifluoroacetic acid or trifluoroacetic anhydride was therefore discontinued, and activating and coupling solutions having the composition acetic anhydride-acetic acid (4:l v/v) were used for all subsequent solution and solid-phase studies. It should be noted that the solubility of a large peptide in the coupling solution should be less of a problem for the solid-phase degradation since the peptide would be distributed as a uniform film on the solid support and its covalent immobilization should minimize aggregation. Thiohydantoin Cleavage in Solution. The rate of cleavage of acetyl-Phe-Leu-2-thiohydantoin by a number of cleavage solutions is shown in Table V. The slow cleavage with 6 and 12 M HCl was similarly observed by Dwulet (1976), and Stark (1968) reported a half-time of 26 min for the cleavage by 10 M HCl of the peptidyl thiohydantoin derived from the oxidized insulin B chain. In contrast, the deacetylation of l-acetyl-2thiohydantoins by concentrated HCl occurs with half-times of about 2 min (Stark, 1968; Darbre, 1977). The reason for this observed difference in thiohydantoin cleavage is not known. The cleavage of acetyl-Phe-Leu-2-thiohydantoin by saturated triethylamine (Dwulet & Gurd, 1979) is rapid (Table V). Preliminary studies on the use of this volatile cleavage reagent in COOH-terminal degradations of short peptides in solution are described below. The cleavage with acetohydroxamate [originally used by Stark (1968)] is rapid and occurs under mild conditions. A cleavage solution of 0.2 M acetohydroxamate in sodium phosphate buffer, pH 8.3, has
3754
BIOCHEMISTRY
MEUTH ET AL.
Table VI: Attachment of Peptides to Carbonyldiimidazole-Activated Porous Glass ~~
amount
L
L A'. i 2 L 6 8 10 12 1L 16 1820 Elution Time (minl FIGURE 2: Identification by HPLC of the amino acid 2-thiohydantoin formed by a single round of the degradation performed in solution on the peptide Gly-L-Leu-L-Tyr. HPLC of (A) Tyr-2-thiohydantoin standard (22 nmol) and (B) ethyl acetate extract of the cleavage solution from the degradation of 800 nmol of Gly-L-Leu-L-Tyr redissolved in 400 pL of methanol. (2.5% of the sample was injected for analysis.)
been used in solution studies and in solid-phase versions of the degradation (see below). COOH- Terminal Solution Degradation. The peptides Gly-L-Leu, DL-Leu-DL-Phe, and Gly-L-Leu-L-Tyr were each subjected to one round of the COOH-terminal solution degradation in order to determine the yield of thiohydantoin. From the degradation of Gly-L-Leu, 100% of the expected Leu-2-thiohydantoin was recovered (by GLC), from the degradation of DL-Leu-DL-Phe, 94% of the expected Phe-2thiohydantoin was obtained (by GLC), and from the degradation of Gly-L-Leu-L-Tyr, 108% of the expected Tyr-2thiohydantoin was recovered (by HPLC, see Figure 2). In the degradation of Gly-L-Leu-L-Tyr, no significant preview (which would have been indicated by the presence of Gly-2thiohydantoin or Leu-2-thiohydantoin in the cleavage solution) was observed. The method provides a rapid procedure for determining the COOH-terminal residue of a peptide. It is limited to those peptides that are soluble in the coupling solution and may be conveniently used for only a single cycle since removal of the cleaving reagents from the peptide (achieved by gel filtration for large peptides; Stark, 1968) is likely to be a time-consuming procedure. The combination of the thiocyanic acid coupling step with a cleavage performed with saturated triethylamine constitutes a method for stepwise degradation using volatile reagents. Preliminary studies on such a procedure have been performed by using model peptides., Attachment of Peptides to Porous Glass. It has been argued that the most useful supports for solid-phase ",-terminal sequencing of proteins have been prepared by the derivatization of glass with aminoalkylsilanes (Laursen & Machleidt, 1980). These supports have also been used for the covalent immobilization of peptides for COOH-terminal sequencing (Williams & Kassell, 1975; Rangarajan & Darbre, 1976). The methods here describe the preparation of a highly substituted aminopropylsilyl glass and the modification and activation of this glass for peptide attachment. The high efficiency of peptide attachment [Table VI; compare Williams & Kassell (19791 results from a number of factors: the high degree of substitution of the glass (typically 0.3-0.6 mmol of reactive group/g of glass), the use of carbonyldiimidazole to give a highly activated glass, the use of anhydrous conditions for coupling peptides to the glass, and the treatment of peptides
peptide
of peptide
Gly-Leu Gly-Leu-Ala Gly-Leu-Tyr Met-enkephalin Me t-enkephalin Kb CB-3b Kb CB-3b glucagon Kb CB-lC
1.O pmol 1.0 pmol 0.8 pmol 1.O pmol 0.75 pmol 0.35 pmol 0.24pmol 90 nmol 0.18 umol
amount of attachactivated ment glass (mg)= (%I 100 100 200 100 200 150 150 150 100
>95 >95 93 57 85 >95 67 76 29
All couplings were performed in 1 mL of dimethylformamide. COOH-terminal CNBr peptide from pygmy sperm whale myoglobin consisting of residues 132-153. NH,-terminal CNBr peptide from pygmy sperm whale myoglobin consisting of residues 155. a
al t
n
A
A
'L
2 L 6 8 10 12 1L 16 18 20 Elution Time iminl
3: Identification by HPLC of the amino acid 2-thiohydantoins formed after sequential degradation of glass-bound Gly-L-Leu-L-Tyr (660 nmol of peptide attached to 175 mg of glass). The cleavage solution of each cycle (4 mL) was collected in 1-mL fractions which were analyzed by HPLC. HPLC of (A) the second 1-mL fraction of the cleavage solution of cycle 1 (the peak eluting at 11.3 min was identified by comparison with standards as 1.74 nmol of Tyr-2thiohydantoin) and (B) the first 1-mL fraction of the cleavage solution of cycle 2 (the peak eluting a t 14.3 min was identified as 2.08 nmol of Leu-2-thiohydantoin). 1% of each sample was injected for analysis. FIGURE
with trifluoroacetic acid to improve their solubilities in the coupling solvent. The improvement in solubility of a peptide in an organic solvent such as dimethylformamide after trifluoroacetic acid treatment has been previously reported (Tarr, 1977; Laursen & Machleidt, 1980). COOH-Terminal Solid-Phase Degradation. The results from experiments with short peptides are shown in Table VI1 and Figure 3. Amino acid analysis of the peptide-glass after sequencing indicates that the degradation generally results in the removal of approximately 90% of the carboxyl-terminal residue. This compares with typical values of 94-98% for the removal of the amino-terminal residue by the Edman degradation (Walsh et al., 1981). The recovery yields of the 2thiohydantoins obtained after each cycle are lower than 90%. The reason for the low recovery of the 2-thiohydantoins observed in the solid-phase degradation is not known. Solvents and reagents used in a degradative cycle have been analyzed by HPLC and by reconversion and amino acid analysis, and thiohydantoins were detected only in the cleavage reagent. The result of performing six cycles of the degradation on the peptide Kb CB-3 is shown in Table VI11 and Figure 4. Identification of the carboxyl-terminal sequence is possible for the first three residues, and tentative assignment of the fourth residue can be made. The sequence agrees with that determined by conventional techniques (Dwulet et al., 1977; M. L.
VOL. 21, NO. 16, 1982
COOH-TERMINAL PEPTIDE SEQUENCING
3755
Table VII: Sequencing of Peptides Attached to Porous Glass from the Carboxyl Terminus
amount of peptide (nmol)
amount of porous glass (mg)
no. of degradative cycles performed
round 1
Gly-Leu Gly-Leu-Ala Gly-Leu-Ala Gly-Leu-Tyr Gly-Leu-Tyr Met-enkephahf
1060 999 1075 680 659 412
100 100 100 176 175 129
1 2 2 1 2 2
Leu45%(GLC) Ala40%(GLC) Ala 63% (HPLC) Tyr75%(HPLC) Tyr 61% (HPLC) Met56%(HPLC)
Leu 39% (HPLC) d
Met-enkephalinc
508
100
4
Met45%(GLC)
Phe63%(GLC)
e
Val-Leu-Ser-Glu-Gly
136
124
4
Gly 75% (HPLC)
Glu 42% (HPLC), Gly 9% (HPLC)
f
peptide
amino acid 2-thiohydantoins identified in cleavage s o h a round 2
composition of peptide attached to glass after sequence analb
round 3
Gly (l.O), Leu (